Quantitative assessment of the quality of tone-mapped images using TMQI-II

Even though the knowledge of tone mapping is continuously developing, many unanswered questions and hypotheses still remains to be solved. On that account, we are quite far away from having an accurate computational model for the compression of high dynamic range images to low dynamic ones. It is; therefore; of particular importance and necessity to conduct experimental subjective evaluations of the computational approaches that incorporate various tone mapping operators to judge and validate them properly. Moreover, the evaluations give us a deeper insight into the inquired area and, besides the evaluation of existing methods they can lead to proposals of new ones. The goal of Image Quality Assessment (IQA) is to computationally predict human perceptions of image quality. In order to improve the quality of the image two factors are considered, they are structural fidelity and statistical naturalness. Then the proposed iterative algorithms are used, so as to improve the quality of the image. Numerical and subjective experiments make evident that the proposed algorithm consistently produces better quality tone mapped images even when the initial images of the iteration are created by the most competitive TMOs.